Although the deinstitutionalization of the seriously mentally ill has been widely criticized, the success of some model community programs shows that community care can reduce relapse and improve patients' quality of life. However, at this point, we do not know which components of programs are responsible for successful outcomes or why these components are effective. This research addresses these issues by proposing and testing a theoretical model which draws connections between specific program elements and multiple outcomes and postulates the mechanisms or intervening variables by which they are related. Program components common to effective model programs include psychiatric treatment, rehabilitation, interpersonal contacts, supervision and guidance, a structure of activities, basic needs, and an approach to service delivery and treatment termed "empowerment". These components are proposed to affect patients; rehospitalizations and objective and subjective quality of life both directly and indirectly, through their effects on factors crucial for the mental status and well-being of chronic patients. Specifically, program components affect outcomes insofar as they increase patients; social support, self-esteem, sense of mastery, and incentives for participation in treatment. Although research on the critical components for outcomes is generally recognized as important, the measures and methods for such research have not been established. thus, in order to test this model, we propose an exploratory study of a model program using measures of discrete program components and methods of coordinating such information from multiple sources developed from intensive observation within this program. We use these measures and methods for a test of the theoretical model on a sample of 70 chronic schizophrenic patients over two points in time to provide preliminary information on predictors of outcome and to refine the model for further, broader-based research. By identifying those components of services which are critical for specific desired outcomes, this research is expected to yield valuable information for the more efficient and effective implementation of community services for the seriously mentally ill.